Spatial distribution patterns of non-stationary extreme precipitation under changing environment in the Yangtze River Delta region

In the current complex changing environment, understanding the non-stationary characteristics of extreme precipitation is increasingly critical for effective flood control and disaster risk management at regional scale. This study examines the Yangtze River Delta （YRD） region using gridded daily precipitation data from 1961 to 2021, along with climate factors. Principal component analysis （PCA） and the generalized extreme value distribution theory （GEV） are employed to analyze how extreme precipitation responds to urbanization. The results show that: ① During periods of rapid urbanization, extreme precipitation in the YRD region intensifies, with short-duration events becoming increasingly uneven in spatial distribution. ② A general upward trend of extreme precipitation is observed across most areas, with more pronounced increases in rapidly urbanizing regions. ③ The monsoon-informed GEV-PCA model performs better in most areas, effectively capturing distributional shifts associated with urban expansion. ④ As return periods increase, the discrepancy in extreme precipitation between the baseline GEV-PCA model and the GEV-PCA-U model （which incorporates impervious surface data） widens, while monthly variability diminishes. These findings highlight that the extreme precipitation is nonstationary and shaped by both atmospheric circulation and urbanization. Regional strategies for flood management and drainage planning should be developed accordingly.